Editor's note

Psychotherapy appears to be effective when compared to naturalistic outcome. While the benefits of psychotherapy may differ across different disorders and different types of psychotherapy may be more effective for certain disorders and less effective for other disorders, repeatedly psychotherapy has shown itself to be more effective than naturalistic outcome. While there are many different types of psychotherapy, they essentially fall into two large classes, the psychodynamic therapies, which includes interpersonal psychotherapy, and the behavioral and cognitive-behavioral therapies. There are also a number of other therapies classified as supportive therapy and experiential therapy. Group and systems, including family, therapy can be subsumed under the above groupings. Currently, there is great attention being paid to both the behavioral (including cognitive-behavioral) and the interpersonal psychotherapies because these seem amenable to study via randomized control methodology. These therapies are also time-limited which facilitates their empirical study because there is a specified time limit or time point at which outcome can be measured. What the essential ingredients are that contribute to a successful therapy remain speculative, but client factors, therapist factors, and the maintenance of a healthy and positive alliance between client and therapist appear to be major influences. Others factors include a well-defined contract, encouraging openness in the patient, and maintaining a focus on current life problems and relationships. There is little evidence to support long-term open-ended therapy no matter what the approach.

In the next four chapters, we will examine biological, psychological, and social factors for the personality disorders. Any of these could be associated with mental disorders: if they make the illness more likely, they are risk factors; if they make the illness less likely, they are protective factors.

In a diathesis-stress model, diatheses are inborn individual differences that influence the vulnerability to mental disorders. The biological risk factors to be examined in this chapter could be diatheses for personality disorders. However, they do not, by themselves, explain their etiology.

The biological factors in mental illness can be measured either through genetic studies, or by the identification of biological markers.

Genetic predispositions to psychiatric disorders are identified by family history methods, by adoption studies, or by twin studies. Family history methods, which determine how frequently a disorder is found in the close relatives of patients, provide information that can only be suggestive of genetic influence, since they cannot separate heredity from environment. On the other hand, adoption and twin studies offer much stronger evidence for biological factors. Adoption studies examine whether the children of parents with psychiatric disorders will develop the same disorder if raised in another family. Twin studies, which determine whether identical twins are more concordant for a disorder than are fraternal twins, are the most common way to measure heritability.

Biological markers are an indirect measurement of heritability. Although they could, in principle, reflect either the causes or the consequences of psychopathology, their presence, particularly in combination with other evidence, points to genetic diatheses in mental disorders (Gottesman, 1991).

To what extent is personality shaped by childhood experiences?

Personality disorders begin early in life, and remit only slowly with time. These characteristics are, of course, intrinsic to their definition. Yet they still demand an explanation. Early onset and chronicity could be accounted for by two alternative theories: (1) the origins of personality disorders lie in temperament; (2) the crucial psychological risk factors for these disorders derive from childhood experiences.

The hypothesis that early experiences, particularly problems in parent-child relationships, can either shape or deform personality, has been taken for granted by generations of theorists. Psychoanalytic theory has explained the consistency of personality over time by assuming a primacy for early experience. The principle is that early learning should have a greater impact than later learning, since it occurs at a time when the child is more dependent on its parents (Millon, 1969). Psychodynamic models all assume that the more severe the pathology, the earlier in life is its origin (Paris, 1983).

In spite of their ubiquity, these assumptions have not been supported by empirical research. One of the most important findings of research in developmental psychopathology is that, by and large, single negative events do not usually cause psychiatric disorders (Rutter & Rutter, 1993). Given a reasonably favorable environment, most children are resilient. In vulnerable children, who lack resilience, stressors have a greater impact. The pathogenic effects of negative events therefore involve interactions with preexisting diatheses.

The present chapter is concerned with methods used to establish whether social factors are involved in the etiology of personality disorders. If the social context should prove to be crucial, important clinical implications would follow. Social factors might help explain why individuals exposed to similar psychological risk factors develop or do not develop personality disorders. The social context might also help explain why certain treatment strategies for personality disorders are effective, while others are not, and why some recover from personality disorders, while others do not.

The social context of mental disorders

Social psychiatry is concerned with the effects of social factors on the causes, course, and treatment of mental illness. However, the difficulty for empirical research in this area is that social risk factors for psychiatric disorders are difficult to measure. There is no practical way to conduct controlled experiments in which the role of social influences can be isolated from other etiological factors. Research in social psychiatry uses indirect methods, and its conclusions inevitably require some degree of inference.

The standard epidemiological methods for establishing etiological relationships are prospective follow-up studies or casecontrol studies. In prospective studies, general community populations, or populations at risk, are followed over a number of years to see which individuals develop a disorder. Casecontrol studies compare patients who have already developed an illness to those without the disorder for the presence of risk factors.

Antisocial personality disorder

Phenomenology

Antisocial personality disorder (ASPD) has been an accepted category of mental illness for two centuries. The idea that there is a form of mental disorder characterized by callousness and criminality is probably universal in all cultures (Murphy, 1976). Many different terms – “moral insanity”, “psychopathy”, “sociopathy” – have been used in the past to describe its phenomenology. Although some writers (e.g., Blackburn, 1988) criticize the construct as a medicalization of “personal deviance”, there is a broad consensus in psychiatry as to its validity.

Since the definitions of antisocial personality disorder in DSM-IV, or of dissocial personality disorder in ICD-10 are similar, we will use them interchangeably. Both systems use primarily behavioral criteria for diagnosis. The criteria in DSM require that there be a pervasive pattern of disregard and violation of the rights of others, as indicated by at least three of the following: criminal actions, deceitfulness, impulsivity, aggressiveness, recklessness, irresponsibility, and lack of remorse. This pattern must have begun before age 15, and be associated with a prior diagnosis of conduct disorder. (The symptoms of conduct disorder are essentially childhood versions of the phenomena seen in adult ASPD.) However, as shown in studies in the United States (Robins et al., 1991), in the United Kingdom (Zoccolillo et al., 1992), and in Australia (Rey et al., 1995) only one third of cases of conduct disorder go on to antisocial personality disorder.

This cluster of disorders is characterized by traits associated with anxiety. When these traits interfere either with the capacity to work or the ability to develop intimate relationships, we can diagnose a personality disorder.

Avoidant personality disorder

Avoidant personality disorder (APD), or its equivalent in ICD, anxious personality disorder, is characterized by a hypersensitivity to rejection that leads to avoidance of intimate relationships. This diagnosis was one of the most commonly made in the international study of personality disorders (Loranger et al., 1994).

The DSM-IV definition of APD requires the presence at least four of the following seven criteria (somewhat paraphrased): avoidance of interpersonal contacts at work; unwillingness to get involved with people unless certain of being liked; restraint in intimate relationships; preoccupation with criticism and rejection; inhibition in new interpersonal situations; view of self as inept, unappealing, or inferior; reluctance to take risks or engage in new activities.

Clinical example

A 32-year-old woman presented at a clinic for evaluation of lifelong social anxiety. Once she felt secure in any interpersonal situation, these difficulties would remit. For example, she was an effective secretary, having worked in the same office for over 10 years. She had also lived with the same female roommate for the last 4 years, and was quite comfortable either at home, or with her family. The patient had not, however, been able to sustain a significant relationship with a man. She had attempted to develop such relationships in her 20s, but had given up after several disappointments.

In the present chapter, we will suggest mechanisms by which social factors could either increase or decrease the risk for developing personality disorders. If the frequency of personality traits varies from one society to another, then the frequency of the disorders associated with these dimensions should vary accordingly. Therefore, in the first section, we will consider to what extent social factors can shape traits. In the second section, we will focus on mechanisms by which social factors could lower the thresholds for traits to develop into disorders.

We can only understand the influence of social factors, on either traits or disorders, in the context of social structures. Social scientists have many constructs to classify these structures, but at this point, we will introduce one basic distinction, which will pervade all the arguments to be made later.

Traditional and modern societies

Social structures can be dichotomized into “traditional” or “modern” types (Lerner, 1958; Inkeles & Smith, 1974). We can best understand this distinction from a historical perspective. Traditional societies are primarily characterized by slow rates of social change and by intergenerational continuity. Modern societies, in contrast, are characterized by rapid social change and by intergenerational discontinuity.

There are, of course, important differences between the traditions of all the societies that might be classified as “traditional”. What we will argue here is that whatever the specific nature of traditions, they provide individuals with a set of predictable expectations.

What is a personality disorder?

Mental disorders produce a wide range of distressing symptoms. Patients may suffer from the profound gloom of depression, the terror of a panic attack, or the disturbing unreality of psychosis. By and large, psychiatric symptoms are experienced as alien and painful.

Personality disorders, in contrast, may or may not cause subjective distress. Their core features are maladaptive patterns of behavior. Some patients report painful inner experiences, but others may not even agree that they have pathology. Some behaviors lead to consequences that make individuals unhappy, while other behaviors, at least in the short run, are more likely to make other people unhappy.

In essence, personality disorders are characterized by inflexible, pervasive, and stable behaviors that cause significant dysfunction in the life of the patient. They begin early in life and are enduring. We can illustrate these basic attributes with two clinical vignettes, describing one patient with distress, and one with almost no distress.

Case 1

A 23-year-old woman was undergoing training for a professional career. In spite of her external success, she thought about suicide nearly every day. In her personal life, she had many unsuccessful love affairs, in which she became overly attached to men who showed insufficient interest in her. When these relationships ended, she would become despondent.

Her problems had begun early in life. At age 13, she had attempted suicide with an overdose of pills. By the time she was 18, she had experimented with a variety of drugs, and was sexually promiscuous.

The previous chapter has proposed a general theory to account for the common factors in the development of any personality disorder, using a model that considers interactions of biological, psychological, and social factors. However, the risks for one type of disorder may not apply to another. We therefore need to examine each diagnostic category separately. Since some disorders have been researched extensively, while others have not been examined in any systematic way, we will allot more space to those about which some degree of empirical data is available.

As reviewed in Chapter 1, the categories of personality disorder described in DSM and ICD-10 have uncertain validity. They overlap with each other, probably due to shared personality dimensions. For this reason, it will be useful to consider them in groups. The next three chapters will review specific categories within the three clusters of disorders described by Axis II of DSM. In addition, we will examine each category separately, in relation to phenomenology, etiology, outcome, and treatment.

Phenomenology of the odd cluster disorders

The personality disorders in the A cluster (schizotypal, schizoid, and paranoid) are associated with unusual thoughts and behaviors, as well as with an inability to establish meaningful interpersonal relationships.

Schizotypal personality disorder

Both DSM and ICD describe a schizotypal disorder, characterized by eccentricity, social deficits, and cognitive distortions. The DSM diagnosis of schizotypal personality disorder requires at least five of the following criteria: ideas of reference, odd beliefs or magical thinking, unusual perceptual experiences, odd thinking and speech, suspiciousness or paranoid ideation, inappropriate or constricted affect, eccentric behavior, lack of close friends, and excessive social anxiety.

We will now apply the guidelines proposed in the previous chapter, and offer a set of recommendations for clinical practice to help patients make better use of their personality traits. The primary goal of this method is improved adaptations.

Before proceeding, two disclaimers must be registered. First, it is not the intention of the author to describe a new type of psychotherapy. Over the last 50 years, many such proposals have appeared, leading to the creation of innumerable “schools” of therapy. Unfortunately, even when these schools bear new labels, when scrutinized, they usually turn out to be variants of existing therapeutic approaches. The model of therapy being proposed here is in no way meant to be unique!

Second, the ideas to be presented in this chapter reflect the author's clinical experience and his theoretical perspective, but remain unsupported by data. At best, they are consistent with the findings of empirical research. No approach to psychotherapy can be considered as valid without systematic research: we need to manualize methods, and then submit them to rigorous clinical trials.

Unfortunately, as reviewed in the previous chapter, there have been no large-scale studies of the efficacy of psychotherapy for personality disorders. Because of the practical difficulties and expense involved in conducting such research, the data at our disposal derive from brief therapies, which may or may not be applicable to the chronic problems presented by personality disordered patients. Clinical guidelines, including the ones to be presented here, can only be provisional.

This chapter will evaluate present methods of treatment for patients with personality disorders. First we will examine the outcome of these disorders, since any claims for successful treatment must be measured against their natural course. Second, we will consider what empirical research can tell us about the effectiveness, or the ineffectiveness, of the most common modalities of therapy. Third, we will attempt to determine which patients are most likely to benefit from treatment. Finally, we will describe certain pitfalls in the therapy of personality disordered patients.

Outcome of the personality disorders

We cannot fully understand the effects of treatment in psychiatric disorders unless we know their outcome. Chronic disorders remit only slowly over time. A gradual improvement in the course of treatment may therefore only reflect natural history.

Personality disorders are, by definition, chronic. We need to examine the empirical literature as to their outcome. But before doing so, we must address three methodological problems. First, outcome research in North America prior to 1980 suffered from an absence of well-defined criteria for diagnosis. Second, not all studies assess patients from multiple points of view; to be comprehensive, research needs to take into account symptoms, levels of functioning in different sectors, as well as whether the criteria for the original personality disorder diagnosis are still present. Third, many outcome studies have been relatively short term (<5 years), while fewer have been truly long term (≥15).

Personality disorders: the history of an idea

In the past, most of the present categories of personality disorder were not considered to be mental illnesses. We need to explain how people with characterological problems came to be seen as meriting psychiatric diagnoses.

Personality disorders are exaggerations of normal personality traits. At some point, these exaggerations produce significant levels of dysfunction, and can therefore be considered pathological. The problem is where to draw the line between normality and pathology.

All medical illnesses lie on a continuum with normality. The determination of what is a “case” is in many respects a social construct (Eisenberg, 1986). If we consider the two examples of personality disorders presented in the introduction, these patients might be considered, in the first case, unwise or unlucky in love, or, in the second case “more bad than mad”. What justifies seeing these people as having mental disorders?

The acceptance of personality disorders as valid diagnoses reflects a change in psychiatric ideology. In order to understand this change, we need to know its historical context.

The classification of personality has a long history that can be traced back to the Greeks (Frances & Widiger, 1986; Tyrer & Ferguson, 1988; Tyrer et al., 1991). A theory describing four temperaments (choleric, sanguine, phlegmatic, and melancholic), associated with the Roman physician Galen (Kagan, 1994), dominated thinking about abnormal personality for many centuries. In fact, if one considers Galen's temperamental types as descriptions, and if one ignores his anachronistic physiological speculations, the four temperaments still have a certain validity (Frances & Widiger, 1986).

The nature and origin of personality traits

Every individual has a set of unique behavioral characteristics, popularly called “personality”. Psychologists define personality traits as consistent patterns of behavior, emotion, and cognition. These characteristics vary greatly between one individual and another. Personality traits can be identified early in life, and are highly stable over time. The broadest characteristics of personality change very little between the ages of 18 and 60 (McCrae & Costa, 1990).

Let us consider an example. Perhaps the most basic of all individual differences in personality is extraversion vs. introversion (McCrae & Costa, 1990; Eysenck, 1991). Extraverts need to be around people, and require a higher level of stimulation. Introverts need more time alone, and require a lower level of stimulation. These differences are rooted in temperament. Extraverts and introverts, if they go on to develop personality pathology, will have different types of disorders.

Both genetic and environmental factors play a role in shaping personality. Rutter (1987) suggested that two factors influence the development of traits: temperament and social learning. Temperament describes those behavioral dispositions present at birth. Observations of newborn- infants show that they differ from each other in how active they are, in how sociable they are, in how easily they get upset, and in how readily they can be calmed down (Kagan, 1994).

Within normal ranges, infantile temperament is not notably continuous with later personality. The only consistent finding of long-term follow-up studies of children with different temperamental dispositions is that a “difficult” temperament, i.e., being easily upset and hard to calm down, makes psychopathology in adulthood more likely (Chess & Thomas, 1990).